refactor: TextToSpeech接口实现

Author: liuruibin

apps/application/serializers/application_serializers.py

@@ -516,7 +516,7 @@ def list(self, with_valid=True):
         @staticmethod
         def reset_application(application: Dict):
             application['multiple_rounds_dialogue'] = True if application.get('dialogue_number') > 0 else False
-            del application['dialogue_number']
+
             if 'dataset_setting' in application:
                 application['dataset_setting'] = {'search_mode': 'embedding', 'no_references_setting': {
                     'status': 'ai_questioning',
@@ -717,14 +717,11 @@ def edit(self, instance: Dict, with_valid=True):
                 self.update_reverse_search_node(instance.get('work_flow'), application_dataset_id_list)
 
             update_keys = ['name', 'desc', 'model_id', 'multiple_rounds_dialogue', 'prologue', 'status',
-                           'dataset_setting', 'model_setting', 'problem_optimization',
+                           'dataset_setting', 'model_setting', 'problem_optimization', 'dialogue_number',
                            'api_key_is_active', 'icon', 'work_flow', 'model_params_setting']
             for update_key in update_keys:
                 if update_key in instance and instance.get(update_key) is not None:
-                    if update_key == 'multiple_rounds_dialogue':
-                        application.__setattr__('dialogue_number', 0 if not instance.get(update_key) else 3)
-                    else:
-                        application.__setattr__(update_key, instance.get(update_key))
+                    application.__setattr__(update_key, instance.get(update_key))
             application.save()
 
             if 'dataset_id_list' in instance:
@@ -823,6 +820,16 @@ def save_other_config(self, data):
 
             application.save()
 
+        def speech_to_text(self, filelist):
+            # todo 找到模型 mp3转text
+            print(self.application_id)
+            print(filelist)
+
+        def text_to_speech(self, text):
+            # todo 找到模型 text转bytes
+            print(self.application_id)
+            print(text)
+
     class ApplicationKeySerializerModel(serializers.ModelSerializer):
         class Meta:
             model = ApplicationApiKey

apps/application/urls.py

@@ -63,5 +63,8 @@
     path(
         'application/<str:application_id>/chat/<chat_id>/chat_record/<str:chat_record_id>/dataset/<str:dataset_id>/document_id/<str:document_id>/improve/<str:paragraph_id>',
         views.ChatView.ChatRecord.Improve.Operate.as_view(),
-        name='')
+        name=''),
+    path('application/<str:application_id>/<str:model_id>/speech_to_text', views.Application.SpeechToText.as_view(), name='application/audio'),
+    path('application/<str:application_id>/<str:model_id>/text_to_speech', views.Application.TextToSpeech.as_view(), name='application/audio'),
+
 ]

apps/application/views/application_views.py

@@ -534,3 +534,35 @@ def get(self, request: Request, current_page: int, page_size: int):
                 ApplicationSerializer.Query(
                     data={**query_params_to_single_dict(request.query_params), 'user_id': request.user.id}).page(
                     current_page, page_size))
+
+    class SpeechToText(APIView):
+        authentication_classes = [TokenAuth]
+
+        @action(methods=['POST'], detail=False)
+        @has_permissions(ViewPermission([RoleConstants.ADMIN, RoleConstants.USER],
+                                        [lambda r, keywords: Permission(group=Group.APPLICATION,
+                                                                        operate=Operate.USE,
+                                                                        dynamic_tag=keywords.get(
+                                                                            'application_id'))],
+                                        compare=CompareConstants.AND))
+        def post(self, request: Request, application_id: str, model_id: str):
+            return result.success(
+                ApplicationSerializer.Operate(
+                    data={'application_id': application_id, 'user_id': request.user.id, 'model_id': model_id})
+                .speech_to_text(request.FILES.getlist('file')[0]))
+
+    class TextToSpeech(APIView):
+        authentication_classes = [TokenAuth]
+
+        @action(methods=['POST'], detail=False)
+        @has_permissions(ViewPermission([RoleConstants.ADMIN, RoleConstants.USER],
+                                        [lambda r, keywords: Permission(group=Group.APPLICATION,
+                                                                        operate=Operate.USE,
+                                                                        dynamic_tag=keywords.get(
+                                                                            'application_id'))],
+                                        compare=CompareConstants.AND))
+        def post(self, request: Request, application_id: str, model_id: str):
+            return result.success(
+                ApplicationSerializer.Operate(
+                    data={'application_id': application_id, 'user_id': request.user.id, 'model_id': model_id})
+                .text_to_speech(request.data.get('text')))